Download Mutational Analysis of Muscle Nicotinic Acetylcholine Receptor

Mutational Analysis of Muscle Nicotinic
Acetylcholine Receptor Subunit Assembly
P a u l B l o u n t a n d J o h n P a u l Merlie
Department of Pharmacology, Washington University School of Medicine, St. Louis, Missouri 63110
Abstract. The structural elements required for normal
hE muscle-type acetylcholine receptor (AChR),t originally isolated from Torpedo electric organ, is the best
characterized member of the family of ligand-gated ion
channels and an ideal subject for studying the structural elements required for normal assembly and expression of multimeric proteins. The AChR is composed of four different but
homologous subunits with multiple transmembrane domains
(Noda et al., 1983) in a stoichiometry of ot2/37~ (Karlin,
1980). The two ot subunits are not juxtaposed, having one
subunit between them on one side and two on the other (Wise
et al., 1981; Kistler et al., 1982; Zingsheim et al., 1982; Bon
et al., 1984). The analysis of predicted protein sequences derived from cDNA clones for each of the subunits of muscle
AChR from several species (Noda et al., 1983; Mishina et al.,
1986), and subunits cloned thus far for the neuronal nicotinic
acetylcholine (Boulter et al., 1986; Goldman et al., 1987),
GABA^ (Schofield et al., 1987), and glycine (Grenningloh et
al., 1987) receptors indicate at least two shared structural features that may lead to common posttranslational modifications
T
1. Abbreviations used in this paper:. AChR, acetylcholine receptor; Bq'x, c~
bungarotoxin.
© The Rockefeller University Press, 0021-9525/90/12/2613/10 $2.00
The Journal of Cell Biology, Volume 111 (No. 6, Pt. 1), Dec. 1990 2613-2622
toxin binding. The unassembled wild-type ot subunit
expressed in fibroblasts is normally degraded with a
tlr2 of 2 h; upon assembly with the 6 subunit, the
degradation rate slows significantly (tt/2 > 13 h). All
mutated c~ subunits retained the capacity to assemble
with a 6 subunit coexpressed in fibroblasts; however,
mutated t~ subunits that were not glycosylated or did
not acquire high affinity toxin binding were rapidly
degraded (tlr2 = 20 rain to 2 h) regardless of whether
or not they assembled with the ~ subunit. Assembly
and rapid degradation of nonglycosylated acetylcholine
receptor (AChR) subunits and subunit complexes were
also observed in tunicamycin-treated BC3H-1 cells, a
mouse musclelike cell line that normally expresses
functional AChR. Hence, rapid degradation may be
one form of regulation assuring that only correctly
processed and assembled subunits accumulate, and ultimately make functional receptors in AChR-expressing
cells.
in this family of proteins. First, 2 cysteines, 14 amino acids
apart, that occur at positions 128 and 142 in the AChR a
subunit, are thought to form a disulfide bridge. Second, a conserved N-linked glycosylation site, located at asparagine 141
in the AChR et subunit, is just before the second of these cysteines.
Previous studies on the kinetics of maturation and assembly of the ot subunit expressed in a muscle cell line, BC3H-1,
suggested that glycosylation and a time-dependent posttranslational modification that confers high affinity tx bungarotoxin
(BTX) binding were required for the efficient assembly and
subsequent expression of the AChR (Merlie et al., 1982;
Merlie and Lindstrom, 1983). Here, we have generated by
site-directed mutagenesis, and transfected into fibroblasts,
several AChR o~subunits that are not glycosylated and/or do
not achieve a mature conformation. We present data consistent with the hypothesis that a disulfide bridge between cysteines 128 and 142 is required for attainment of a mature
conformation of a subunit as assayed by BTX binding. In addition, by using the association of subunits in transfected
fibroblasts as an assay for assembly, we show that mutated
tx subunits retain the capacity to assemble with ~ subunit to
2613
Downloaded from jcb.rupress.org on August 3, 2017
maturation and assembly of the nicotinic acetylcholine
receptor a subunit were investigated by expression of
mutated subunits in transfected fibroblasts. Normally,
the wild-type ct subunit acquires high affinity ol bungarotoxin binding in a time-dependent manner; however, mutation of the 128 and/or 142 cysteines to either serine or alanine, as well as deletion of the entire
14 amino acids in this region abolished all detectable
high affinity binding. Nonglycosylated subunits that
had a serine to glycine mutation in the consensus sequence also did not efficiently attain high affinity binding to toxin. In contrast, mutation of the proline at
position 136 to glycine or alanine, or a double mutation of the cysteines at position 192 and 193 to serines
had no effect on the acquisition of high affinity toxin
binding. These data suggest that a disulfide bridge between cysteines 128 and 142 and oligosaccharide addition at asparagine 141 are required for the normal
maturation of ot subunit as assayed by high affinity
form o~5 complexes; however, relative to the wild-type or6
complex, complexes made from mutated ~ subunits that
were not glycosylated or did not acquire high affinity BTX
binding were quickly degraded. Thus, rapid degradation of
immature subunits and subunit complexes may be a mechanism by which incorrectly assembled receptor complexes are
edited to preclude surface expression of nonfunctional
AChR.
Materials and Methods
Cell Growth and Labeling
Growth conditions for BC3H-I cells (Merlie and Sebbane, 1981; Merle and
Lindstrom, 1983) and QT-6 cells (Moscovici et al., 1977; Blount and Merlie, 1988) as well as transfection, selection and maintenance of QT-6 clones
expressing ~ (Blount and Merlie, 1988), c~ and % and a and c5 subunits
(Blount and Merlie, 1989) of the mouse AChR have been described previously; all subsequent clones have been derived and maintained similarly.
Pulse labeling with [35S]methionine was performed at a final specific activity of"~6 Ci/mmol for long pulses or 800 Ci/mmol for 5-rain pulses (Blount
and Merlie, 1988). For cells that were chased, the media was removed subsequent to the pulse, and was replaced by conditioned media supplemented
with l-mM unlabeled methionine.
All constructions are in M13 shuttle vectors (Blount and Merlie, 1988,
1989). Single-strand preparations were made and point mutations were induced (Burke and Olson, 1986) using mutating oligonucleotides with 10 bp
flanking the mutating base(s). After transformation, M13 plaques were
screened using differential hybridization of the mutating oligonucleotide.
The oligonucleotides used to make the deletion mutants (128-142) and the
142C~A mutant had 18-19 bp flanking the deleted bases; these mutants
were screened with oligonucleotides with 10 bp flanking the deletion. The
normal sequence from 127 to 144 were changed to Tyr Asn Gly Ser Met
and Tyr Ser Set Ser Met in the glycosylated and not glycosylated deletion
mutants, respectively. The 192 + 193 CC~SS mutation was made with an
oligonucleotide with two mismatches. All other double mutants weremade
by sequential mutations. All mutations were confirmed by single stranded
sequencing of the mutated area using an internal oligonucleotide primer and
Sequenase obtained from United States Biochemical (Cleveland, OH).
Replicative form preparations purified through two cesium chloride gradients were made from these constructions for calcium phosphate transfection. Transient transfections were harvested 2 d after transfection; stable
clones were selected and screened for subunit expression (Blount and Merlie,
1988). The cell line expressing the ~ and ~5subunits was stably transfected
with the/~, 7, and ~ subunits; expression of the 7 subunit has not been detected by coimmunoprecipitation of subunits or immunofluorescence.
Binding Assay
The binding of t25I-BTX to membranes prepared from fibroblast cells has
been previously described (Blount and Merlie, 1988) with modifications
(Blount and Merlie, 1989). To determine if mutant ~6 complexes contained
carbamyleholine binding sites, an inhibition of the initial rate of 12SI-BTX
binding assay (Blount and Merlie, 1988, 1989) was used.
Results
Bungarotoxin Binding and Assembly of
Subunit Mutants
The ~ subunit of the AChR acquires high affinity BTX binding in a time-dependent manner before (Merlie and Lindstrom, 1983) and independent of (Blount and Merlie, 1988)
assembly with other subunits; however, the covalent modification(s) required for the maturation of the ot subunit is unknown. Merlie and Lindstrom (1983) speculated that the formation of a disulfide bridge may be the modification leading
to a mature c~ subunit. Only four cysteines exist in the extracellular region of the AChR ot subunit. Two cysteines that
make a rare adjacent cysteine disulfide bridge (Kao et al.,
1984) are located at position 192 and 193 of the muscle ot
subunit and are conserved in all ot subunits of the muscletype and neuronal nicotinic receptors (Fig. 1 A). Two additional cysteines are located at position 128 and 142 of the
A
r7
~~M1/~--~-~M3
156-179
M4e/-~tracellular
..... 4 = - - - = t 1 - ~ ...... ~ " . . . . . . . . X .......................... t ~ t ~ ; ~ i ~
e
Of O~..O-~136,,0
t
Immunoprecipitations and Related Methods
......... T//'~C128 "',"" C142- S 1T)143'--"/~ .........
Labeled cells were washed twice with PBS, 300/~M PMSF at 4°C, scraped
from the plates, and pelleted by centrifugation for 10 s in a microfuge. The
cell pellet from a single 10-era dish was extracted with 1 ml PBS, 1-2%
Triton X-100, 200 t~M leupeptin, and 0.2 U/nil ofc~2 macroglobulin for 3-5
min, then the cells were centrifuged for 5 min in a microfuge and the supernatant collected. Immunoprecipitations were performed as previously described (Merlie and Sebbane, 1981; Merlie and Lindstrom, 1983) with
modifications (Blount and Merlie, 1988). The antibodies used were the c~
subunit-specific monoclonal antibodies mAb61 (Tzartos et al., 1981) and
mAb210 (Rathman et al., 1986), a rabbit polyclonal anti-BTX antibody
used to precipitate cc subunit that was prebound to BTX (toxin antitoxin)
(Merlie and Sebbane, 1981; Merlie and Lindstrom, 1983), the ~ subunitspecific mAb148 (Gullick and Lindstrom, 1983), and the 6 subunit-specific
mAb88B (Froehner et al., 1983). After immunoprecipitation, Staphylococcus aureus pellets were resuspended in sample buffer and subjected to SDSPAGE on a 10% acrylamide, 0.27 % N,N'-bis-methylene acrylamide gel and
buffer system previously described (Laemmli, 1970). The gels were processed for fluorography; band intensity was proportional to radioactivity
and exposure time (Laskey and Mills, 1975). The percentages of association
Figure I. Structure of the muscle nicotinic AChR c~ subunit. A
shows some structural features of the u subunit. These include the
four highly conserved membrane spanning domains labeled M1 to
M4 and a putative loop structure formed by a disulfide bridge,
depicted here as a circle, at amino acid positions 128-142. Also
shown are the two cysteines located at position 192 and 193 that
are unique to the c~ subunit; these cysteines are conserved in some
subunit clones isolated for neuronal AChRs. An epitope between
amino acid positions 156 and 179 has been mapped to the cytoplasmic side of the protein by antibody binding to Torpedo vesicles
(Criado et al., 1985; Pedersen and Cohen, 1990a) and is depicted
here as a "V" dipping into the cytoplasmic domain. The putative
loop structure boxed in A is shown in greater detail in B. The proline approximately half way between the two cysteines and the
glycosylation site just before the second cysteine are highly conserved among a family of ligand-gated channels.
The Journal of Cell Biology, Volume I l l , 1990
2614
Downloaded from jcb.rupress.org on August 3, 2017
Construction of a Subunit Mutations and
Molecular Techniques
between subtmits were calculated by measuringthe fluorogramintensities
with a scanning densitometer (LKB Instruments, Inc., Gaithersburg, MD)
and adjusting for immunoprecipitation effieiencies using t25I-BTX-labeled
BC3H-1 AChR as an internal standard (Blount and Merlie, 1988).
made as described in experimental procedures; stably transfected fibroblasts eoexpressing mutated c~and wild-type 6 subunits of the AChR
were labeled with [35S]methionine for 1 h, harvested, and extracted as described in Materials and Methods. The mutated ¢x subunits were
immunoprecipitated with the c~-specific monoclonal antibodies mAb61 and rnAb210. Toxin antitoxin was used to immunoprecipitate c~
subunit that had acquired high affinity BTX binding (c~ax). The 6 subunit specific monoclonal antibody mAb88B was used to immunoprecipitate 6 subunit, and c~ subunit associated with 6 subunit. Precipitates were analyzed by SDS-PAGE and the fluorograms are shown.
Amino acid changes resulting from the mutations are designated by position, wild-type amino acid~mutant amino acid; each panel is
labeled at the bottom. The 6 subunit is marked by a bracket because of its heterogeneity in SDS-PAGE, and the ¢x subunit is labeled with
a tick. As seen in 136P--'A, 6 subunit is sometimes observed as a doublet. We suspect this to be a partially glycosylated state of the 6
subunit; note the previously observed (Blount and Merlie, 1988) nonglycosylated tx subunit in the same panel. The ratio of a precipitated
by BTX to total ~xvaried between experiments; no consistent difference was observed in the efficiency of ¢xaxformation between the wildtype cz subunit, 136P--'(G or A), or 192 + 193 CC~SS. Sequences for the deletion mutants A128-142, glycosylated and not glycosylated,
are presented in Materials and Methods. 136 shows an immunoprecipitation of a cell line expressing only the/3, or expressing the/3 and
6 (/36) AChR subunits. The lane BNS shows immunoprecipitation of a cell line expressing only the 13subunlt with the/3-specific mAb148.
The/36 cell line was immunoprecipitated with mAb148 and mAb88B; the lane NS is an immunopreeipitation where only the second antibody
for mAb88B and S. aureus were added. The mutated c~, the wild-type 6, and the nonglycosylated mutated c~ subunits (c~c) are labeled.
Note that because the bands observed in the BNS lane are derived from a cell line that does not produce the 6 subunit, the band eomigrating
with 6 in the BNS and 148 lanes must be a nonspecific protein. This discrete nonspecifie band has been previously observed (Blount and
Merlie, 1989; Blount et al., 1990), and is also labeled in the 192 + 193 CC'-"SS mutant.
c~ subunit, and are conserved in all ligand-gated channel subunits cloned thus far with the exception of glutarnate receptor
(Hollmann et al., 1989). These two cysteines are thought
to form a disulfide bridge with a proline approximately midway, and an N-linked glycosylation site one amino acid before the second cysteine (Fig. 1, A and B). To determine
if these conserved elements are required for the normal
maturation of the AChR ot subunit, a series of mutations
were made in the ~x subunit by site directed mutagenesis. Expression vectors encoding mutated proteins were transiently
or stably transfected into QT-6 fibroblasts. These cells were
then metabolically labeled for 1 h with [3~S]methionine,
harvested, immnnoprecipitated, analyzed by SDS-PAGE,
and visualized by fluorography. To assay for high affinity
BTX binding, an antitoxin antibody was used to immunoprecipitate BTX prebound to ot subunit (toxin antitoxin) (Merlie
and Sebbane, 1982; Merlie and Lindstrom, 1983). As shown
Blount and Merlie Mutational Analysis of AChR Assembly
in Fig. 2, mutation of the proline at position 136 to alanine
or mutation of the cysteines at positions 192 and 193 to serines had no detectable effect on the expression of oLsubunits
acquiring high affinity BTX binding in this assay, suggesting
that the posttranslational modification leading to the formation of a high affinity BTX binding site is not dependent upon
the 136 proline and cannot be the disulfide bridging of one
or both of the cysteines located at 192 or 193. In contrast,
high affinity BTX binding was significantly reduced when
the glycosylation consensus sequence was mutated (143S--"
G), and no BTX binding was detected when cysteine at position 142 was changed to alanine. Deletion of the amino acids
between position 128 and 142 in the ot subunit resulted in no
detectable high affinity BTX binding whether or not the
glycosylation site was preserved.
From results of several mutations reviewed in Table I,
some generalizations can be made. First, glycosylation is
2615
Downloaded from jcb.rupress.org on August 3, 2017
Figure 2. Immunoprecipitation of AChR subunits from fibroblast cells expressing mutated ¢x and wild-type/t subunits. Mutations were
The Journal of Cell Biology, Volume 111, 1990
2616
Mutation*
Wild type
192 + 193 CC~SS
Glycosylation
143S-"G
143S--,'G + 148T~N
Loop mutations
Cysteines
128C~S
128C~A
142C~S
142C~A
128C~S + 142C~S
Proline
136P~G
136P~A
Deletions
A128-142 glycosylated
A128-142 not glycosylated
BTX binding~t
a/~ association§
+(6/6)
+(7/7)
+(2/2)
+(2/2)
- (7/8)11
-
+ (5/5)
+
-(3/3)
-(5/5)
-(3/3)
-(5/5)
-(6/6)
+
+(4/4)
+
+(1/1)
+(3/3)
+ (10/10)
+(5/5)
+ (6/6)
+(3/3)
-(6/6)
-(3/3)
+(5/5)
+(2/2)
* Mutations were made as described in Materials and Methods. Amino acid
changes resulting from these mutations are designated by position, wild-type
amino acid~mutant amino acid. The sequence of the deletion mutants are
presented in Materials and Methods.
¢ High affinity BTX binding was assayed by toxin antitoxin immunoprecipitation of extracts from cells pulse labeled with 13SS]methionine for 1 h. Precipitates were analyzed by SDS-PAGE and fluorography as in Fig. 1 and
densitomelric scanning. In some cases results were confirmed with a ~z~I-BTX
binding assay previously described (Blount and Mealie, 1988). Positive results
(+) ranged from '~15 to 40% of total t~ subunit with signals 3--40-fold above
background determined by nonspecific immunopreeipitation. Negative results
were less than or equal to background. All results were originally observed in
transient transfections. Where applicable, the number of clones expressing the
mutation showing the same properties/the total number of clones expressing
the mutation are shown in brackets.
Association of oc and 6 subunits was assayed by coimmunoprecipitation of
the c~subunit by the ~ subunit specific mAb88B from extracts of cells pulse labeled with [35S]methionine for 1 h. Precipitates were analyzed by fluorography and densitometric scanning. Positive results (+) ranged from '~7 to 35%
of total et subunit. All results were originally observed in transient transfections of the mutant c~ subunits into a QT-6 clonal cell line transfected with and
stably expressing the 5 subuult. Stably transfected cell lines expressing the mutant a subunits and the wild-type 6 subunit were made by cotransfection of the
two constructs and selection and screening of clones as described in Materials
and Methods. Where applicable, the number of clones expressing the mutation
showing the same properties/the total number of clones expressing the mutation are shown in brackets.
II In one high expressing stably ttansfected clone expressing this mutation, a
long fluorographic exposure suggested that <2 % of the ~ subunit was precipitated with toxin antitoxin as determined by densimetric scanning.
Downloaded from jcb.rupress.org on August 3, 2017
necessary for efficient acquisition of BTX binding; however,
in long exposures of fluorograms from high expressing
clones, ,,o2 % of the nonglycosylated t~ subunit made was detected using toxin antitoxin suggesting that glycosylation is
not an absolute requirement. A mutation that reconstituted
a glycosylation consensus sequence seven amino acids down
stream from the normal glycosylation site (143S---G +
14STUN) resulted in an o~ subunit with electrophoretic
migration identical to that of the glycosylated form, but did
not restore the efficiency of acquisition of BTX binding; thus,
the requirement for glycosylation is site specific. Second,
cysteines at positions 128 and 142, but not at 192 and 193,
are required for acquisition of high affinity BTX binding. Deletion of the amino acids between 128 and 142 (A 128-142,
glycosylated or not glycosylated) confirmed that preservation of secondary structure in this region is a requirement for
normal expression of high affinity BTX binding. Finally, mutation of proline at position 136 had no effect, suggesting that
this amino acid is not critical for acquisition of high affinity
binding of toxin. All of these data are consistent with the hypothesis that a disulfide bridge between cysteines 128 and
142 and glycosylation at asparagine 141 are required for the
normal maturation of ct subunit of the AChR.
Previously, we demonstrated that membranes derived
from stably transfected fibroblasts expressing the wild-type
a subunit had a K ~ of 9.5 x 10-H M for ~25I-BTXin a 22-h
binding assay (Blount and Merlie, 1988). In art attempt to
determine the extent of alteration of the binding properties
of ot subunit mutants, we measured the apparent affinity of
125I-BTX to membrane preparations from some of the fibroblast cell lines expressing mutant c~ subunits. The 192 + 193
CC--'SS mutant had a K,pp for J25I-BTX binding of 9.2 x
10-~° M, an affinity approximately 10-fold less than the
wild type. However, no specific binding was observed for the
128C~S + 142C-*S mutant in this assay at concentrations
of 125I-BTXup t o 1 x 10-s M. Thus, mutations of the cysteines at position 128 and 142 had dramatic effects (>1,000fold) on affinity of BTX for AChR ot subunit. In contrast, mutations of cysteines at position 192 and 193 reduced affinity
of BTX by 10-fold.
As an assay for assembly, shuttle vector constructions encoding mutant oL subunits were transiently transfected into,
and expressed in a fibroblast cell line stably expressing only
wild-type ~ subunit; alternatively, stably transfected fibroblasts expressing both a and 6 subunits were isolated. The
coimmunoprecipitation of subunits using subunit specific
antibodies has been shown to be a sensitive and specific assay
for assembly of these subunits (Blount and Merlie, 1990).
However, the 6 subunit is often difficult to detect because it
migrates on SDS-PAGE as a heterogeneous band, comigrates with a nonspecifically precipitated protein, and is extremely sensitive to proteolysis. Therefore, the coimmunoprecipitation of ~ with 6 specific mAb88B (Froehner et al.,
1983), a more sensitive and reliable assay than coimmunoprecipitation of 6 with ct specific antibodies, was used to routinely detect or/6 association. One caveat to these experiments is that high levels of expression or nonstoichiometric
expression of subunits in transfected fibroblasts may lead to
nonspecific subunit association. However, characterization
of several clones expressing varying stoichiometries of o~and
subunits as well as consistent observations made in transiently transfected fibroblasts decreases the likelihood of artifactual subunit association. Although many of the mutated
t~ subunits did not achieve a mature conformation as assayed
by their ability to bind BTX with high affinity, they all assembled with 6 subunit in a coimmunoprecipitation assay (Fig.
2 and Table I). Coimmunoprecipitation of subunits was also
tested in a cell line stably coexpressing the/3 and 6 subunits
(Fig. 2). Although the/3 subunit shares 40% identity with
et subunit, and shares many of the same secondary structural
features, 13 did not efficiently assemble with the ~ subunit.
These data suggest that assembly of/3 with ~ subunit may be
a late event in normal assembly of AChR subunits. Hence,
we conclude that structural domains present in ot that are required for efficient assembly with 6 subunit are not present
in the homologous/3 subunit. Furthermore, we suggest that
although glycosylation and a disulfide bridge between cysteines 128 and 142 are required for normal maturation of
Table L The Effect of Amino Acid Mutation and Deletion
on the Ability of the tr Subunit to Bind BTX
and Associate with $ Subunit
Figure 3. Association of a and
a subunit, they are not necessary for the assembly of tx and
subunits.
The ot~ and ot~¢complexes form high affinity binding sites
for small agonists and competitive antagonists that do not
bind the unassembled a subunit (Blount and Merlie, 1989).
In addition, a recent affinity labeling study suggests that the
binding site for these ligands may be at the interface of or3,
and cx~ subunits (Pedersen and Cohen, 1990b). However, the
structural requirements for the formation of this binding site
are not known. Therefore, to determine if the 192 and 193
cysteines or the proline at position 136 were requirements for
the formation of an agonist binding site, we tested the ability
of a high concentration (1 mM) of carbamylcholine, a small
agonist, to inhibit Iz~I-BTX binding in membranes prepared
from clones expressing mutant ot and wild-type ~ subunits.
These studies demonstrated that in fibroblasts coexpressing
the 192 + 193 CC--'SS or the 136P--'A mutant with fi subunit greater than 50% of the normal 12sI-BTXbinding could
be inhibited by carbamylcholine, a value similar to clones
co-expressing the wild-type tx and 3' or ct and 5 subunits
(Blount and Merlie, 1989); many of the 125I-BTX binding
sites that were not inhibited by carbamylcholine are presumably unassembled a subunits that bind tzq-BTX but not carbamylcholine (Blount and Merlie, 1988, 1989). Therefore,
the cysteines at positions 192 and 193 and the proline at position 136 in the AChR et subunit are not critical for formation
of a carbamylcholine binding site.
Assembly of Nonglycosylated a and ~ Subunits in
Tunicamycin-treated Fibroblasts
Experiments using the N-linked glycosylation inhibitor,
tunicamycin, suggested that glycosylation of subunits was
important for normal maturation and assembly of mouse
AChR subunits in BC3H-1 cells (Merlie et al., 1982). However, as seen in Fig. 2, a mutation that inhibits glycosyla-
Blount and Merlie MutationalAnalysisof AChRAssembly
tion in the ot subunit did not prevent assembly of ot with
subunit in stably transfected fibroblasts. To determine if
tunicamycin was preventing assembly of AChR subunits by
a mechanism independent of inhibition of ot subunit glycosylation, we incubated a wild-type or8 fibroblast cell line
(Blount and Merlie, 1989, 1990) in the presence or absence
of tunicamycin, pulse labeled the cells for 1 h with [3sS]methionine, and immunoprecipitated with subunit specific antibodies. As seen in the fluorogram presented in Fig. 3, tunicamycin treatment of or8 cells increased the mobility of ot and
subunits on SDS-PAGE and substantially decreased the
amount of tx subunit precipitated by toxin antitoxin, but did
not prevent the assembly of tx subunit with 6. These data are
consistent with the charactermaUon of the ot subunit glycosylation mutation presented in Fig. 2, but are in apparent contradiction with the data obtained in BC3H-1 cells (Merlie et
al., 1982).
•
.
~.
.
Assembly and Degradation of Normal and
Nonglycosylated AChR Subunits in Untreated and
Tunicamycin-treated BC3H-1 Cells
The half-life for degradation of unassembled AChR subunits
in BC3H-1 cells has been determined to be only 15-30 min
(Merlie and Lindstrom, 1983). Therefore, it seemed possible that nonglycosylated subunits expressed in tunicamycintreated BC3H-1 cells did assemble, but that assembled products were degraded as rapidly as unassembled subunits. In
an attempt to detect subunits that assembled but degraded rapidly, we repeated the study of tunicamycin treated BC3H-1
cells using very short [35S]methionine pulse-labeling and
chase times. In this experiment, coimmunoprecipitation of
ot with either B or 8 subunit specific antibodies (not available
for the previous study) was used as an assay for assembly.
BC3H-1 cells were pulse labeled with [3~S]methionine for
2617
Downloaded from jcb.rupress.org on August 3, 2017
8 subunits expressed in tunicamycin-treated fibroblasts. Confluent 60-min dishes of the tx/t
cell line were incubated in the
absence (Control)or presence
of 1.5/zg/ml Tunicamycin for
10 h. The ceils were labeled
with [35S]methionine for 1 h,
and then harvested and extracted as described in Materials
and Methods. The ot subunit
was immunoprecipitated with
the tx-specific monoclonal antibodies mAbtl and mAb210.
Toxin antitoxin was used to
immunoprecipitate ~t subunit
that had acquired high affinity
BTX binding (otax).The ~ subunit-specific monoclonal antibody mAb88B was used to
immunoprecipitate/~ subunit,
and ct subunit associated with
the ~ subunit. Precipitates were
analyzed by SDS-PAGE and the fluorograms are shown. The lane NS is an immunoprecipitation where only the second antibody and
S. aureus were added. Note the nonspecific band labeled with asterisks. The wild-type c~ and 5, and the nonglycosylated tx and 8 (trNc
and 3~6) subunits are labeled.
the first 80 min as assayed by the coimmunoprecipitation of
a subunit by a B subunit-specific antibody. After a 5-min
pulse only a small amount of ot is assembled with/3 subunit
and the amount of assembly increases with increasing chase
times. In contrast, in tunicamycin-treated BC3H-1 cells, no
increase of assembled nonglycosylated o~ subunit was observed, and the nonglycosylated /~ subunit was quickly
degraded with a tt/2 of ,x,43 min (Fig. 4 B). However, in
long fluorographic exposures, a small amount of assembly
was detected in tunicamycin-treated cells (Fig. 4 C). Nonglycosylated but assembled ct subunit was degraded with a
tl/2 of •16 rain, essentially the same as the unassembled,
nonglycosylated ot subunit (Table II). Although we obtained
similar results in this experiment using coimmunoprecipitation of ot with ~ subunit-specific antibody, the efficiency of
the/3 subunit-specific antibody, mAb148 (Gullick and Lindstrom, 1983), and the resistance of the/3 subunlt to proteolysis made the coimmunoprecipitation of c~ with the/3 subunit
Table II. Half-Life for Degradation of Unassembled
and Assembled Subunits
Half-life for degradation
In BC3H-1 Cells*
Unassembled
subunit
Surface AChR
Nonglycosylated
unassembled
c~ subunit
Nonglycosylated
t~ assembled with
/3 subunit
Figure 4. Kinetics of assembly and/or degradation o f subunits in
normal and tunicamycin-treated BC3H-1 cells. Confluent 10-cm
dishes o f the BC3H-1 cell line were incubated in the absence (A)
or presence of 1.5/~g/ml tunicamycin (B and C) for 10 h. The cells
were labeled with [35S]methionine for 5 m i n and chased with conditioned media supplemented with I m M methionine for the times
(in minutes) indicated (Chase Time). The cells were then harvested
and extracted as described in Materials and Methods. The/3 subunit
was immunoprecipitated with the/3-specific monoclonal antibody
mAb148. The precipitates were analyzed by SDS-PAGE, and the
fluorograms are shown. A is a 6-d fluorographic exposure; B and
C are from the identical gel, 5- and 27-d fluorographic exposures,
respectively. The/~, the nonglycosylated/3 (~N6), and the c o i m m u noprecipitating a and nonglycosylated et subunit (c~N~) are labeled.
In transfected fibroblasts
Wild-type ~ subunit
Wild-type a subunit
192 + 193 CC-"SS
a subunit
136P-~G c~ subunit
143S~G ct subunit
1 2 8 C ~ S + 142C~S
subunit
1 4 2 C ~ A ct subunit
A128-142 glycosylated
c~ subunit
15-30 min
8 h
13 min
16 min
Unassembled
2 h
2 h
Assembled with
Not applicable
> 13 h
ND
ND
17 rain
> 13 h
> 13 h
22 min
1.2 h
2.0 h
1.0 h
1.8 h
1.5 h
1.7 h
only 5 min; some cells were then harvested while others
were chased with conditioned media for increasing times before harvesting. Fig. 4 A shows a fluorogram from untreated
BC3H-1 cells showing the normal course of assembly over
All half-lives for degradation were calculated from densitometric scans of appropriate fluorographic exposures from pulse-chase labeling experiments. The
half-life for degradation of unassembled wild-type c¢ subunit in transfected
fibroblasts was previously published (Blount and Merlie, 1988) and has been
confirmed in three independent clones. The half-life for degradation of wildtype <5subunlt was obtained from a cell line producing only the 6 subunit. Mutations were made by site-directed mutagenesis as described in Materials and
Methods. Amino acid changes resulting from these mutations are designated
by position, wild-type amino acid--,-mutant amino acid. The half-life for degradation of unassembled, was calculated from densitometric scans of fluorograms of ~ subunit pulse-chase labeled and immunoprecipitated with mAb61;
ND, not determined. The half-life for degradation of the unassembled 128C~
S + 142C-~S c~subunit was not significantly different between a'clone producing both the mutated c~ and ~ subunit and a clone producing only this mutated
subunlt. The half-lives for degradation Of wild-type and mutated a subunits
assembled with 6 were calculated from densitometric scans of fluorographs
with ~ coimmunoprecipitated with the &specific antibody mAb88B.
* The half-lives for degradation of unassembled c~ subunit (Merlie and Lindstrom, 1983) and surface AChR (Patrick et al., 1977) in BC3H-1 cells were
previously published. The half-lives for degradation of the nonglycosylated c~
subunlt in tunicamycin-treated BC3H-I cells were calculated from densitometric scans of appropriate fluorographic exposures from the experiment shown
in Fig. 4.
The Journal of Cell Biology, Volume 111, 1990
2618
Downloaded from jcb.rupress.org on August 3, 2017
Subunit
a more sensitive assay. In addition, data demonstrating that
ot and/~ subunits associate only inefficiently in transfected
fibroblasts (Blount and Merlie, 1989) suggests that the association of these subunits may be a late event in assembly,
and perhaps, a better assay for more complete and normal
AChR assembly. Thus, the nonglycosylated ot subunit expressed in tunicamycin-treated BC3H-1 cells has the potential to assemble with other subunits, but the assembled products are rapidly degraded.
Kinetics of Assembly and Degradation of
Mutant and Wild-type ~ Subunit Complexes in
Transfected Fibroblasts
tated and wild-type c~//i cell lines. Confluent 10-cm dishes of cell
lines expressing mutated or wild-type czand/i subunits were labeled
with [35S]methionine for 5 min and chased with conditioned media
supplemented with 1 mM methionine for the times indicated
(Chase llme in minutes, and TIME(HOURS)in C). The cells were
then harvested and extracted as described in Materials and
Methods. The/i subunit was immunoprecipitated with the/t-specific monoclonal antibody mAb88B and the precipitates were analyzed by SDS-PAGE. Fluorograms are shown in A and B. A shows
an experiment from a cell line coexpressing 143S--'G c~and/i subunit, and B is from a cell line coexpressing 128C-'*S + 142C-'*S
tx and/i subunit. The/i and the coimmunoprecipitating nonglycosylated ct (OING)and mutated tx subunit are labeled. C shows quantitafive analysis, by densitometric scanning, of assembled ot in a cell line
coexpressing wild-type ~x and/i subunits. The error bars in C are
SEM of three independent experiments; data for each experiment
were normalized to the maximum assembled c~ subunit.
Discussion
Expression of functional multimeric transmembrane recep-
Blount and Merlie MutationalAnalysisof AChRAssembly
2619
Downloaded from jcb.rupress.org on August 3, 2017
Figure 5. Kinetics of assembly and degradation of subunits in mu-
As demonstrated in Fig. 4, tunicamycin inhibition of N-linked
glycosylation in the BC3H-1 cell line did not prevent assembly of nonglycosylated subunits; however, assembled products were rapidly degraded. To determine ifa similar process
occurs in transfected fibroblasts coexpressing a mutant, nonglycosylated, ot subunit with wild-type 5 subunit, we labeled
these cells with [35S]methionine for 5 min and immunoprecipitated using a 5 subunit specific antibody. As seen in Fig.
5 A, nonglycosylated a subunit coimmunoprecipitated with
subunit was degraded with a t~/2of 22 min, a degradation
rate similar to the unassembled nonglycosylated tx subunit
(Table II). Similarly, the degradation rates of ol subunits with
mutations at cysteines 128 and 142 or a deletion in this region did not decrease upon association with ~ subunit (Fig.
5 B and Table ID. In contrast, wild-type cz and 5 subunits assembled in fibroblasts in a time-dependent manner and, as
previously described (Blount and Merlie, 1990), the resulting complex had a half-life of 13 h, 6-10-fold greater than
either the unassembled ct or 5 subunits (Fig. 5 C and Table
II). Note that a small amount of assembly of wild-type ot with
6 was observed at the earliest time point measured (Fig. 5 C,
5-min pulse with no chase). This early assembly may be the
same as that observed at the earliest time points in Fig. 5,
A and B; however, in contrast to cells expressing wild-type
a, accumulation of assembled complexes from these mutant
ot subunits was not observed.
Mutations that do not prevent acquisition of high affinity
BTX binding could have subtle effects on kinetics of ot
subunit maturation and assembly. However, experiments
similar to those presented in Fig. 5 demonstrated that in
fibroblast cell lines co-expressing wild-type ~ with either the
136P--'G or the 192 + 193 C C ~ S S ot subunit, t~ subunit acquired high affinity BTX binding and assembled with 5 subunit with kinetics and efficiency indistinguishable from wildtype ot subunit (not shown). In addition, the ix6 complex
formed in these cell lines had a degradation rate similar to
the wild type (Table II).
A summary of our data on degradation rates of assembled
and unassembled subunits is presented in Table II. Collectively, these data suggest that glycosylated and cysteine 128
and 142 disulfide bridged tx subunits that have assembled
into a heterologous tx~5complex accumulate, while unassembled subunits and or6 complexes composed of tx subunits that
are not glycosylated or have not made a normal 128-142
disulfide bridge are degraded.
ever, the deletion mutants suggest that the secondary structure formed by these amino acids is critical for a mature conformation. Analysis of double cysteine mutants (192 + 193
CC-~SS and 128C-*S + 142 C ~ S ) support the conclusion
that our observations are due to lack of disulfide bridging,
not formation of improper disulfide linkages or the presence
of free sulfhydryls. Because mutation of either one or both
of the cysteines at position 128 and 142 leads to absence of
detectable BTX binding in our assay, and because these
amino acids have not been implicated in agonist or BTX
binding (as 192 and 193 cysteines have), our data strongly
suggest that a disulfide bridge between cysteines at position
128 and 142 is required for acquisition of high affinity BTX
binding and a mature conformation of A C h R , subunit.
The maturation of the o~subunit is a slow event; maximum
levels of otrx are not achieved until several minutes after o~
subunit synthesis. Possible explanations for this delay include (a) a slow isomerization of the proline at position 136,
or, (b) slow isomerization of disulfide bonds among the 4
cysteines at positions 128, 142, 192, and 193 (Fig. 1 A) until
the correct disulfide configuration is achieved. Several previous studies have suggested that either proline or disulfide
bond isomerization can be rate limiting for normal protein
folding (for review see Fischer and Schmid, 1990). If isomerization of proline 136 or disulfide bond isomerization is
rate limiting, mutation of proline 136 or a double mutation
of cysteines at position 192 and 193 should significantly increase the rate of acquisition of high affinity BTX binding.
Our data, however, indicate that ot subunit mutated at
136P-*G, and the double mutant o~ 192 + 193 CC-*SS, acquire high affinity BTX binding with kinetics and efficiency
indistinguishable from wild-type ~t subunit. Hence, neither
isomerization of the proline at position 136 nor incorrect intramolecular disulfide bridging can account for the timedependent nature of ot subunit maturation.
Merlie et al. (1982) demonstrated that tunicamycin treatment of BC3H-1 cells decreased Otrx formation and AChR
assembly, but not ct subunit synthesis. When it was discovered that ot subunit could acquire high affinity BTX binding
subsequent to glycosylation but before assembly with other
subunits (Merlie and Lindstrom, 1983), a sequential model
for processing and assembly of AChR subunits was proposed. In this model, ct subunit is translated and cotranslationally glycosylated, acquires high affinity BTX binding in
a time-dependent manner, then is assembled with other
subunits to form the AChR. The time-dependent nature of
formation of O~vxseemed sufficient to suggest this as rate
limiting for assembly. However, because experimental intervention in the normal process of ot subunit maturation was
not possible, the evidence that otr~ was required for subunit
assembly relied entirely on the effect of tunicamycin on
BC3H-1 cells and on kinetic studies. And indeed, this model
was recently challenged when Xenopus oocytes injected with
messenger RNA of Torpedo ix,/~, % and ~ AChR subunits
were treated with tunicamycin, and assembly of nonglycosylated subunits was observed (Sumikawa and Miledi,
1989). The results presented here for mouse AChR subunits
expressed in fibroblasts confirm Sumikawa and Miledi's observations. In addition, we demonstrate that although immature t~ subunit can assemble, the assembled products are
short lived when compared to the correctly assembled products. Because Xenopus oocytes have a relatively slow rate of
The Journalof Cell Biology,Volume111, 1990
2620
Downloaded from jcb.rupress.org on August 3, 2017
tors requires assembly of subunits into a correct heterosubunit complex. However, the requirements for and regulation of subunit assembly are not well understood. Because
of our current knowledge of the structure of muscle-type
AChRs, we can use this molecule as a model for studying requirements for normal subunit association. Here we have
used the assembly of AChR o~ and ~ subunits in transfected
fibroblasts (Blount et al., 1990) to analyze the effects of sitedirected mutations of the a subunit in an attempt to identify
structural elements required for normal maturation and assembly of subunits.
The interpretation of our results depends upon our using
BTX binding as an assay for ct subunit maturation. Merlie
and Lindstrom (1983) previously demonstrated that AChR tx
subunit acquired high affinity B-I'Xbinding in a time-dependent
manner. Subsequently, we demonstrated that in stably transfected fibroblasts, ot subunit could acquire high affinity BTX
binding in the absence of other subunits (Blount and Merlie,
1988). These data suggested a covalent modification was required for a subunit maturation as assayed by BTX binding.
The data presented here implicate ot subunit cysteines 128
and 142 in this maturation. However, interpretation of effects
of mutation is not straightforward. The change of a single
amino acid may cause loss of drug binding by interfering
with folding (as we believe is the case for mutations of cysteine 128 and 142), or, alternatively, mutations may directly
influence the drug binding site. Several studies have suggested that an tx subunit domain to which BTX (W'dson et
al., 1985; Neumann et al., 1986; Barakas et al., 1987; Wilson and Lentz, 1988; Radding et al., 1988) and competitive
antagonists (Kao et al., 1984; Dennis, et al., 1988) bind is
located at or near the 192 and 193 cysteines. However, a
study in which mutated TorpedoAChR a subunits were coexpressed with wild-type/3, % and 6 in Xenopus oocytes suggested that cysteines 192 and 193 were not requirements for
BTX binding (Mishina et al., 1985). The authors found that
when cysteine at position 192 or 193 was mutated to serine,
although no agonist gated ion channels were detected on the
cell surface, a significant amount of carbamylcholineinhibitable BTX binding sites reached the plasma membrane. These data suggested that the 192 and 193 cysteines
were not required for expression of agonist or BTX binding
sites, but were required for channel gating. In contrast, when
cysteine at positior~ 128 or 142 was mutated to serine, neither
ligand-gated channels nor high affinity BTX binding sites
were detected on the surface of the oocytes. Because only
BTX binding sites that reached the oocyte plasma membrane
were monitored, these results were inconclusive in determining if the cysteines at 128 and 142 were necessary for formarion of oLsubunit with high affinity BTX binding sites, or,
alternatively, if these amino acids were required for the assembly of a with other subunits or transport of assembled
AChR to the cell surface. Consistent with the results of Mishina et al., 1985, we found that mutation of cysteines 192 and
193 to serines did not interfere with acquisition of high
affinity BTX binding, assembly with 6 subunit, or formation
of a carbamylcholine binding site. In addition, our data indicate that cysteines at position 128 and 142 are required for
expression of any detectable high affinity BTX binding sites.
The observation that mutation of proline at position 136 does
not inhibit ctr~ formation confirms that not all amino acids
in this region are important for ot subunit maturation; how-
We would like to thank Drs. Stanley Froehner and Jon Lindstrom for the
use of their antibodies, and Dr. J. Majors for making available RSV Neo
plasmid DNA. We would also like to thank Despina Ghement for technical
assistance in growth of the BC3H-I cell line.
P. Blount was supported by National Research Service Award 2 T32
GM 07805. This work was also supported by funds from the Senator Jacob
Javits Center of Excellence in the Neuroscienees and research grants from
the National Institute of Health and the Muscular Dystrophy Associations
of America.
Received for publication 5 July 1990 and in revised form 22 August 1990.
References
Barkas, T., A. Mauron, B. Roth, C. Allied, S. J. Tzartos, and M. Ballivet.
1987. Mapping the main immunogenic region and toxin binding site of the
nicotinic acetyleholine receptor. Science (Wash. DC). 235:77-80.
Blount, P., and J. P. Medic. 1988. Native folding of an acetylcholine receptor
a subunit expressed in the absence of other subunits. J. Biol. Chem.
263:1072-1080.
Blount, P., and J. P. Merlie. 1989. Molecular basis of the two nonequivalent
ligand binding sites of the muscle nicotinic acetylcholine receptor. Neuron.
3:349-357:
Blount, P., M. M. Smith, and J. P. Merlie. 1990. Assembly intermediates of
the mouse muscle nicotinic acetylcholine receptor in stably transfected
fibroblasts. J. Cell Biol. 111:2601-2611.
Bon, F., E. Lebron, J. Gomel, R. V. Rapenbusch, J. Cartand, J.-L. Popot, and
J.-P. Changeux. 1984. Image analysis of the heavy form of the acetylcholine
reo~ptor from Torpedo marmorata. J. Mol. Biol. 176:205-237.
Boulter, J., K. Evans, D. Goldman, G. Martin, D. Treco, S. Heinemann, and
J. Patrick. 1986. Isolation of a eDNA clone coding for a possible neuronal
nicotinic acetylcholine receptor ot-subunit. Nature (Load.). 319:368-374.
Burke, D. T., and M. V. Olson. 1986. Oligodeexynucleotide-directed mutagenesis of Eseherichia coil and yeast by simple cotransformation of the
primer template. DNA (NY). 5:325-332.
Criado, M., S. Hochschwender, V. Sarin, J. L. Fox, and J. Lundstrom. 1985.
Blount and Merlie Mutational Analysis of AChR Assembly
Evidence for unpredicted transmembrane domains in acetylcholine receptor
subunits. Proc. Natl. Acad. Sci. USA. 82:2004-2008.
Dennis, M., J. Giraudat, F. Kotzyba-Hibert, M. Goeldner, C. Hirth, J.-y.
Chang, C. Lazure, M. Chrefien, and J.-P. Changeux. 1988. Amino acids
of the Torpedo marmorata acetylcholine receptor ¢xsubunit labeled by a photoaflinity ligand for acetylcholine binding site. Biochemistry. 27:2346-2357.
Fischer, G., and F. X. Schmid. 1990. The mechanismof protein folding. Implications of in vitro refolding models for de novo protein folding and translocation in the cells. Biochemistry. 29:2205-2212.
Froehner, S. C., K. Douville, S. KUnk, andW. J. Clup. 1983. Monoclonal antibodies to cytoplasmic domains of the acntylcholine receptor. J. Biol. Chem.
258:7112-7120.
Goldman, D., E. Deneris, W. Luyten, A. Kochhar, J. Patrick, and S.
Heinemann. 1987. Members of a nicotinic acetylcholine receptor gene family are expressed in different regions of the mammalian central nervous system. Cell. 48:965-973.
Grenningloh, G., A. Rienitz, B. Schmitt, C. Methfessel, M. Zensen, K. Beyreuther, E. D. Guedelfinger, and H. Betz. 1987. The strychnine-binding
subunit of the glycine receptor shows homology with nicotinic acetylcholine
receptors. Nature (Load.). 328:215-220.
Gullick, W. J., and J. M. Lindstrom. 1983. Mapping the binding of monoclonal
antibodies to the acetyicholine receptor. Biochemistry. 22:3312-3320.
Hollmann, M., A. O'Sbea-Greenfield, S. W. Rogers, and S. Heinemann. 1989.
Cloning by functional expression of a member of the glutamate receptor family. Nature (Load.). 342:643-648.
Kao, P. N., A. J. Dwork, R. -R. J. Kaldany, M. S. Silver, J. Widernan, S.
Stein, and A. Karlin. 1984. Identification of the ¢x subunit half-cystine
specifically labeled by an affinity reagent for the acetylcholine receptor binding site. J. Biol. Chem. 259:11662-11665.
Karlin, A. 1980. Molecular properties of nicotinic acetylcheline receptors. In
Cell Surface and Neuronal Function, C. W. Cotman, editor. Elsevier/North
Holland Press, New York. 191-260.
Kistler, J., R. M. Stroud, M. W. Klyrnkowsky, R. A. Lalancett, and R. H. Fairclough. 1982. Structure and function of an acetylcholine receptor. Biophys.
J. 37:371-383.
Kobayashi, S., and H. Aoshima. 1986. Time course of the induction of acetylcholine receptors in Xenopus oocytes injected with mRNA from Electrophorus electricus electroplax. Dev. Brain Res. 24:211-216.
Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of
the head of bacteriophage of 1"4. Nature (Load.). 227:680-685.
Laskey, R. A., and A. D. Mills. 1975. Quantitative film detection of 3H and
14C in polyacrylamide gels by flurography. Eur. J. Biochem. 56:335-341.
Medic, J. P., and J. Lindstrom. 1983. Assembly in vivo of mouse muscle acetylcholine receptor: identification of an a subunit species that may be an assembly intermediate. Cell. 34:747-757.
Merlie, J. P. and R. Sebbane. 1981. Acetyicholine receptor subunits transit a
precursor pool before acquiring c~-bungarotoxin binding activity..1". Biol.
Chem. 256:3605-3608.
Merlie, J. P., R. Sebbane, S. Tzartos, and J. Lindstrom. 1982. Inhibition of
glycosylation with tunicamycin blocks assembly of newly synthesized acetylcholine receptor subunits in muscle cells. J. Biol. Chem. 257:2694-2701.
Mishina, M., T. Tobimatsu, K. Imoto, K. Tanaka, Y. Fujita, K. Fukuda, M.
Kurasaki, H. Takahashi, Y. Morimnto, T. Hirose, S. Inayama, T. Takahashi, M. Kuno, and S. Numa. 1985. Location of functional regions of acetylcholine receptor c~-subunit by site-directed mutagenesis. Nature (Load.).
313:364-369.
Mishina, M., T. Takal, K. Imoto, M. Noda, T. Takahashi, S. Numa, C. Methfesset, and B. Sakmann. 1986. Molecular distinction between fetal and adult
forms of muscle acetylcholine receptor. Nature (Load.). 321:406-411.
Moscovici, C., M. G. Moscovici, H. Jimenez, M. M. C. Lai, M. J. Hayman,
and P. K. Vogt. 1977. Continuous tissue culture cell lines derived from
chemically induced tumors of Japanese quaff. Cell. 11:95-103.
Neumarm, D., D. Barchan, A. Safran, J. M. Gershoni, and S. Fucks. 1986.
Mapping of the c~bungarotoxin binding site within the o~subunit of the acetylcholine receptor. Proc. Natl. Acad. Sci. USA. 83:3008-3011.
Noda, M., H. Takahashi, T. Tanabe, M. Toyosato, S. Kikyotani, Y. Furutani,
T. Hirose, H. Takashima, S. Inayama, T. Miyota, and S. Numa. 1983.
Structural homology of Torpedo californiea acetylcholine receptor subunits.
Nature (Load.). 302:528-532.
Patrick, J., J. McMillan, H. Wolfson, and J. C. O'Brien. 1977. Acetylcholine
receptor metabolism in a nonfusing muscle cell line. J. Biol. Chem.
252:2143-2153.
Pedersen, S. E., and J. B. Cohen. 1990a. Identification of a cytoplasmic region
of the Torpedo nicotinic acetylcholine receptor tx subunit by epitope mapping. J. Biol. Chem. 265:569-581.
Pedersen, S. E., and J. B. Cohen. 1990b. d-Tubocurarine binding sites are located at ~x-.yand ct-/~subunit interfaces of the nicotinic acetylcholine receptor.
Proc. Nan. Acad. Sci. USA. 87:2785-2789.
Radding, W., P. W. R. Corfield, L. S. Levinson, G. A. Hashim, and B. W.
Low. 1988. ix-Toxin binding to acetylcholine receptor ct 179-191 peptides:
intrinsic fluorescence studies. FEBS (Fed. Fur. Biochem. Soc.) Left.
231:212-216.
Ratnam, M., P. B. Sargent, V. Sarin, J. L. Fox, D. Le Nguyen, J. River, M.
Criado, and J. Lindstrom. 1986. Location of antigenic determinants on primary sequences of subunits of nicotinic acetylcholine receptor by peptide
2621
Downloaded from jcb.rupress.org on August 3, 2017
AChR expression (Kobayashi and Aoshima, 1986), and,
most likely a slow rate of degradation of all unassembled and
assembled subuait species, it seems probable that even relatively rapidly degraded subunits are more easily observed in
the oocyte expression system than the BC3H-1 or transfected
fibroblast cell lines. Thus, our data present a resolution to
an apparent paradox of the effects of tunicamycin on assembly of AChR subunits. Moreover, our data suggest that correctly processed and assembled subunit complexes are recognized by cellular transport or degradation machinery in a
manner different from incorrectly processed and assembled
complexes or unassembled subunits, thereby leading to the
selective accumulation of only correct assembly intermediates.
In conclusion, by studying individual and combinations of
subunits expressed in fibroblasts, we have gained insight into
subunit maturation and regulation of assembly of the AChR.
Mutational analysis suggests that acquisition of high affinity
BTX binding requires the formation of a disulfide bridge between cysteines 128 and 142 in ~x subunit. Because this processing step is time-dependent and relatively inefficient, it
remains a possible site of regulation for AChR expression.
The assembly of CXTxwith other subunits and surface expression of AChR appear to be specific and highly regulated.
Previous work has suggested that discrete assembly intermediates are formed in and confined to the endoplasmic
reticulum when o~and 3' or c~ and fi subunits are coexpressed
in fibroblasts (Blount et al., 1990). Mutational analysis now
suggests that subunits lacking N-linked glycosylation or a
disulfide bridge between cysteines 128 and 142 can assemble, but these incorrect assembly intermediates are degraded,
thus leading to accumulation, and ultimately, expression of
only correctly processed AChR.
mapping. Biochemistry. 25:2621-2632.
Schofield, P. R., M. G. Darlison, N. Fnjita, D. R. Burt, F. A. Stephenson,
H. Rodriguez, L. M. Rhee, J. Ramachandren, V. Reale, T. A. Glencorse,
P. H. Seeburg, and E. A. Barnard. 1987. Sequence and functional expression
of the GABA^ receptor shows a ligand-gated receptor super-family. Nature
(Lond.). 328:221-227.
Sumikawa, K., and R. Mfledi. 1989. Assembly and N-glycosylation of all ACh
receptor subunits are required for their efficient insertion into plasma membranes. Mol. Brain. Res. 5:183-192.
Tzartos, S. J., D. E. Rand, B. L. Einarson, andJ. M. Lindstrom. 1981. Mapping of surface structures of electrophorous acetylcholine receptor using
monoclonal antibodies. J. Biol. Chem. 256:8635-8645.
Wilson, P. T., and T. L . Lentz. 1988.' Binding of c~-bungarotoxin to synthetic
peptides corresponding to residues 173-204 of the a subunit of Torpedo,
calf, and human acetylcholine receptor and restoration of high affinity binding by sodium dodecyl sulfate. Biochemistry. 27:6667-6674.
Wilson, P. T., T. L. Lentz, and E. Hawrot. 1985. Determination of the primary
amino acid sequence specifying the c~-bungarotoxin binding site on the a
subunit of the acetylcholine receptor from Torpedo californica. Proc. Natl.
Acad. Sci. USA. 82:8790-8794.
Wise, D. S., J. Wall, and A. Karlin. 1981. Relative locations of the/~ and 8
chains of the acetylcholine receptor determined by electron microscopy of
isolated receptor trimer. J. Biol. Chem. 256:12624-12627.
Zingsheim, H. P., F. F. Barrantes, J. Frank, W. Hanicke, and D. -Ch. Neugebauer. 1982. Direct structural localization of two toxin-recognition sites on
an ACh receptor protein. Nature (Lond.). 299:81-84.
Downloaded from jcb.rupress.org on August 3, 2017
The Journal of Cell Biology, Volume 11 I, 1990
2622